Binary Molecular Property Prediction on MoleculeNet (test)
72BBBP AUCGraphFPCPF
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| GraphFPCPFcontrastive pretraining (C)=true, predictive pretraining (P)=true, fragment encoders (F)=true, alpha=0.12023.10 | 72 | 74 | 63.9 | 63.6 | 84.7 | 75.4 | 78 | 80.5 | 4.56 | 74.01 | |
| GraphFP-JTCFragmentation Strategy=Junction Tree [18]2023.10 | 71.5 | 75.2 | 63.6 | 62 | 77.7 | 76 | 75.6 | 79.7 | 6.13 | 72.66 | |
| GraphFPCcontrastive pretraining (C)=true2023.10 | 71.5 | 75.5 | 63.8 | 61.4 | 78.6 | 77.2 | 76.3 | 78.2 | 5.5 | 72.81 | |
| JOAOv2Pretraining Strategy=JOAOv22023.10 | 71.4 | 74.3 | 63.2 | 60.5 | 81 | 73.7 | 77.5 | 75.5 | 8.94 | 72.14 | |
| GraphFPCPcontrastive pretraining (C)=true, predictive pretraining (P)=true, alpha=0.32023.10 | 71.3 | 75.5 | 64.7 | 61.3 | 73.7 | 76.6 | 76.3 | 81.3 | 5.19 | 72.59 | |
| JOAOPretraining Strategy=JOAO2023.10 | 70.2 | 75 | 62.9 | 60 | 81.3 | 71.7 | 76.7 | 77.3 | 9.56 | 71.89 | |
| GraphFP-JTCFFragmentation Strategy=Junction Tree [18], Includes Fragment Encoder=true2023.10 | 70.2 | 72.7 | 62.5 | 59.3 | 75.9 | 73.9 | 73 | 74.2 | 13.56 | 70.21 | |
| GraphFPCFcontrastive pretraining (C)=true, fragment encoders (F)=true2023.10 | 70.1 | 74.3 | 65.3 | 64.7 | 87.7 | 74.5 | 76.1 | 77.1 | 7.25 | 73.73 | |
| G-ContextualPretraining Strategy=G-Contextual2023.10 | 69.9 | 75 | 62.8 | 58.7 | 60.6 | 72.1 | 76.3 | 79.3 | 11.88 | 69.34 | |
| GraphCLPretraining Strategy=GraphCL2023.10 | 69.7 | 73.9 | 62.4 | 60.5 | 76 | 69.8 | 78.5 | 75.4 | 12.13 | 70.78 | |
| MGSSLPretraining Strategy=MGSSL2023.10 | 68.9 | 74.9 | 63.3 | 57.7 | 67.5 | 73.2 | 75.7 | 82.1 | 10.94 | 70.41 | |
| GraphMVPPretraining Strategy=GraphMVP2023.10 | 68.5 | 74.5 | 62.7 | 62.3 | 79 | 75 | 74.8 | 76.8 | 10 | 71.7 | |
| GraphFPPpredictive pretraining (P)=true2023.10 | 68.2 | 76 | 63.2 | 59.3 | 53.8 | 74.5 | 76.7 | 80.7 | 9.5 | 69.05 | |
| ContextPredPretraining Strategy=Context Prediction2023.10 | 68 | 75.7 | 63.9 | 60.9 | 65.9 | 75.8 | 77.3 | 79.6 | 7.56 | 70.89 | |
| GraphLoGPretraining Strategy=GraphLoG2023.10 | 67.8 | 75.1 | 62.4 | 59.5 | 65.3 | 73.6 | 73.7 | 80.2 | 12.56 | 69.7 | |
| G-MotifPretraining Strategy=G-Motif2023.10 | 66.9 | 73.6 | 62.3 | 61 | 77.7 | 73 | 73.8 | 73 | 14.25 | 70.16 | |
| GPT-GNNPretraining Strategy=GPT-GNN2023.10 | 64.5 | 74.9 | 62.5 | 58.1 | 58.3 | 75.9 | 65.2 | 77.9 | 13.63 | 67.16 | |
| AttrMaskingPretraining Strategy=Attribute Masking2023.10 | 64.3 | 76.7 | 64.2 | 61 | 71.8 | 74.7 | 77.2 | 79.3 | 7.88 | 71.15 |